| |
SM7UZI > BPQ 15.08.99 22:17l 264 Lines 11469 Bytes #-9923 (0) @ EU
BID : SM7UZI6389
Read: GUEST IN3HRR
Subj: Re: BPQ Obsolete
Path: DB0AAB<DB0KFB<DB0CZ<DB0GE<DB0ZDF<DB0SRS<DB0SIF<DB0FHK<DB0HSK<DB0SHG<
DB0SM<DB0PDF<DB0CL<DB0HES<OZ7BOX<OZ5BBS<OZ4PAC<OZ4BBS<OZ4BOX<OZ3BOS<
OZ3PAC<OZ6BBS<SK7IQ
Sent: 990815/1331Z @:SK7IQ.MALMOE.M.SWE.EU #:47317 [Malmö] $:SM7UZI6389 FBB7.00
From: SM7UZI@SK7IQ.MALMOE.M.SWE.EU
To : BPQ@EU
sm3lks wrote:
>But I have not seen any other message, originating in Sweden, that support you.
Supporting me?
It would be nice if people gave it a thought and took a look at some other node
softwares and became a bit more interested in the future of packet radio.
Some debating about how to make the packet network better would be nice!
>What is true, or not, is a relative thing.
More or less!
>What I say i true, someone else says it is wrong.
>And they can both be right.
>An example:
>It's 10 degrees centigrade. Is it cold ?
>That depends what temperature the person is used to.
It is could below zero!
>So you can't tell the whole world the truth, it might not be true everywhere.
>>There is no need for insulting your fellow swedish apr users!
>
>I'm sorry if you got insultet by my statement.
>But honest, when did I insult you?
You didn't insult me.
>When I said that only you says that BPQ is crap ?
When you try to use the swedish packet users uninterest in this matter as some
argument in the debate.
>>All people are not interested in this kind of rather stupid discussions!
>Why do you keep starting this "stupid discussions" then ?
I hope that some of those reading them will be interested in building a better
a packet network.
Most of the replyes are stupid, most of what I write have been written before,
but some times is something of interest written!
Some new thoughts some serious replyes from those who know something about
what they write etc.
>>This is not a discussion, leading to some sort of developement of the packet
>>network or much very good. Maybe will some peole learn something, it is why I
>>still am answearing and so on.
>OK, a good reason, but educate us instead of saing that we don't know
>anything and your way is best.
>You just say a thing, but not explaining so people can understand.
I try to explain what I mean!
A problem is that packet radio is rather advanced and most users don't know
very much about it, maybe not even enough to realise what this is all about.
It is the same reason for why people don't know how to set their tnc
parameters and there for cause alot of problems for their self and their
fellow hams.
Some of those reading this are node operator sysops etc and some of them maybe
have a little more packet knowledge.
But how shall that be sloved?
>Example: Just saing that BPQ has a bad packetdriver isn't enough.
>Everyone doesn't know what a packetdriver is.
They can always ask!
>An explanation should avoid a lot of misunderstanding, and replies.
What should I do?
Include some basic packet review in every bulletin?
>>>I have a working node with a 386SX33, 4 MB RAM, 60 MB HD, FBB v7.00g,
>>>BPQ v4.08a, USCC>4-card (4 ax25-ports).
>>Use tfpcx or flexnet, they are both better packet drivers.
>>Give up the node, both speed and memory is better used by the bbs.
>>But that computer is really not enough for both a node and fbb7, not even
fbb7 >>you shuld get some more ram.
>An exellent example of bad teaching.
maybe!
>As you don't know anything about the actual performance of this computer,
>I fail to see how you can make that statement.
I know the performance of a comptuer like that, have used one for packet for a
while.
I know that 4mrg ram is not enough for fbb7!
It can use up to 16meg ram, so 24meg of ram for best performance.
Hd usage is affecting the node, so a big hd cache is important, espscially if
you don't use scsi hds.
So 24 meg will give you 16meg for fbb, 4meg hd cache and the rest for dos etc!
If you don't want more ram had I used fbb5.15 which requires less of the
computer instead.
The performance of tfpcx is better than bpq, since bpq lack some important
features, but evern better is flexnet!
But for running flexnet and fbb in that computer is it probably to slow!
I woul have got another computer, and used the 386 är flexnet node and the
other as bbs.
But also the conifiguration, hd cache software used, dos used etc affects the
performance.
By loading drivers in the right way will you get more ram free for fbb and the
packet driver to use.
Using a better memmanager then microsofts emm386 make it very much faster.
>You can't possible know how this configuration works.
>Have you seen it in action ?
Very similair setup!
I also know what the problems caused bya s to slow computer is worse when
using baycom modems and uscc modems then if you use drsi cards ot tncs.
>And no, it is not slow.
Compared to how a flexnet node in that computer connected to a bbs computer
with fbb is it.
You will get atleast 20% better throughput, probablly much more.
Especially on user ports, where the number of active stations change from time
to time.
You haven't compared with any other setup!
>>>In what way ?
>>>(Sorry, I forgot: Things manufactured yesterday is automatically obsolete)
>>It seems lit you are not realy interested in how to improve the performance of
>>youy node, the apcket network etc.
>>More like you se this like it is your duty to defent the bpq software.
>
>Im am really interested in improving the network, but I'm realistic.
Try flexnet!
The better packet driver really imporves things, especially on user ports!
>I can't do changes in my corner, and automatically think that everyone is
>about to do the same changes.
So?
Some one has to begin, and to setup a net/rom link via a flexnet network is
possible so you will not destroy the network.
Only improve it!
>Changing to FlexNet can actually be a degrade of the network, as it has
>non-standard responses, the routing is not compateble, etc..
No, the net/rom users set up a link via the flexnet part of the network, the
network users can stil connect every other net/rom node.
But if they want to connect a flexnet node, must they connect to the closest
flexnet node and then use the flexnet routing.
If thenetnodes would be used instead woulf it have been even better.
>In one place where they have a FlexNet-node, the network actually hung for
>extensive periodes because a non-standard resonse from FlexNet.
Don't know what you mean?
Could you please explain, I'm very interested!
Flexnet follows ax25 and works with every other packet driver doing that!
Bpq doesn't really don't follow, that causes protocol errors
some times, but also between bpq nodes etc. But that happens seldom.
If it should be flexnets tries to use the flexnet protocol can you turn
that test of for very link to a bpq node.
>To get the most out of FlexNet, it should be used everywhere.
Yes!
But even when not used every where will it make ting work better!
Only the adaptive parameters of the packet driver make it much better, it aslo
makes it impossible to misconfigura the packet driver.
>This is often an impossible task, as many nodes is unattended and remote,
>and that makes a replacement from NET/ROM-tnc to FlexNet-computer a VERY
>bad thing if you want a network that is stable.
No, rmnc/flexnet is very good for unatrended stations, much better working
than any tnc! More reliable and rather cheap.
Pc/flexnet is not the only alternative.
>(Computers do crash you know, NET/ROM-tnc's has NO problems there)
>That makes a NET/ROM-compatible node better than FlexNet in any place of
>the network.
No, gues you didn't know about tha rmnc hardware!
But rmnc flexnet is much more reliable than any tnc!
>>>They are cheap, and they can do the job with processor-capacity in reserve.
>>Sometimes, but not always.
>>Why do you think computers get faster and faster?
>
>Theuy are getting faster by two reasons:
>1. The companies that makes them want more money.
That they want to sell new computers is one and the major reason I guess.
>2. Computergames etc..., programs with a graphic frontend demands fast
> computers. (AKA windows-based programs)
People want to use more demanding software, or in the case of windows is
forced to.!
>But when you comes to packetradio, you don't need any graphichal frontend,
>it has no big arithmetic operations to perform.
When it comes to nodes does a better routing, header compression etc require
more computer power than a older node with less good routing, no header
compression adaptive parameters etc.
>You don't need a state of the art computer.
No, but the 8088 is not useable for moder pc based nodes!
You need a 386 or 486!
386 is the oldes ps worth using!
>A 486DX33 has almost never more than 10% usage of processorcapacity,
>running a packet-node and BBS in the same computer.
It is good for that.
But how about when the hd is used, how much processor time is needed for that?
But also te hd speed, amount of ram for hd cache, bus speed etc affects the
speed.
But if would be just fin for a pc/flexnet node with yout uscc card!
>For unattended nodes, just using a computer is a bad thing.
>Computers should only be used in attended stations.
Can work fine when using dos or linux.
The more modern pc computers is not like the old days, they are made to be
suitable as servers, bbses etc.
That is why you not should use a 286 etc even if they are fast enough.
A 386 or 486 at least, that aslo makes it impossibel to user better nodes than
bpq.
>
>>>I use a single-processor-computer, and I have not discovered any loss in
>>>the node-performance. (Linux + linux-node)
>>Teh you ahve a fast computer, low speed packet modems, much ram for hd cache
>>etc.
>>But for many people is it much easier to get two old 386, 486 computers to use
>>as node and bbs, than one faster comptuer for both.
>>Not very strange.
>
>I'm using a 486DX66. Thats not fast according to you.
You must have missunder stood med!
For low speed packet is it fast enough.
Would be just fine for a node, bbs, cluster etc.
>It is an old one, you just said so yourself.
Yes is is to slow for most usage, but for packt is it still useble.
I beliave that there are many 386 or 486 computers around that can't be used
for much else than packet, printer servers etc.
They woukd be fins as pc/flexnet nodes.
>>>As a matter of fact, my node (in an obsolete single-processor-computer,
>>>running a multitasking OS) has a less problems than the other in my area.
>>If it is a good node is there never any problems, it shall run for years woith
>>out any breaks! Like most rmnc/flexnet noced!
>>Your node is not like that!
>
>How do you know that ?
>And what node is not like that ?, my Linux-node or the 386-BPQ-node ?
At least if you use fbb and bpq is the system not very stable.
>BPQ-node:
>It has run for a lot of years with just 2 crashes, HD-crashes that could
>just as easy happened if we were running FlexNet.
>The computer has never crashed because the use of BPQ.
No, why should it!
>Linux-node:
>More than 2 years of uninterupted 24h/day usage. No problems.
Nice as it should be!
What node are you using?
.a&$a. .a&$a. .a&$$$$$$$$$$&a. .a&$a. Anders Lageras
$$$$$$ $$$$$$ $$$$$$$' `$$$$$$ $$$$$$ Lund Sweden JO65PQ
$' `$$ $$$$$$ `ý$$$$$s,s$$$$$ý $$$$$$ AARS Member No. 81
$s,s$$ $$$$$$ a$$$$$$ý $$' `$ Packet:
73 de SM7-$$$$$$ $$$$$$ a$$$$$$ý $$s,s$ SM7UZI@SK7IQ.MALMOE.M.SWE.EU
$$$$$$ $$' `$ a$$$$$$ý $$$$$$ E-mail: sm7uzi@qsl.net
$$$$$$a$$s,s$ a$$$' `$$$$$$&a. $$$$$$ ICQ#: 5161379
$$$$$$$$$$$$$ $$$$s,s$$$$$$$$$ $$$$$$ Homepage:
`ý&$$$$$$$$ý' `ý&$$$$$$$$$$$ý' `ý&$ý' http://zap.to/sm7uzi/
o==|=====-
Read previous mail | Read next mail
| |
